That's an optimistic take, or maybe I've become cycnical (regarding human nature). A class based analysis of the bullshitting verbal virtuoso and the role it can play in silencing experts, for example, could point to "positive" outcomes for a subset who field these machines. Regarding the Jinn, just fyi, some of them are asserted to be Muslims. Take the one capable jinnie who time-travelled in service of Solomon so he could impress the Queen of Sheba .. ;)
Sure, but how far can it go? complaints about "the MSM" gained traction after the Bush administration fed false information to the New York Times, which printed it credulously and eagerly. It took exactly one misstep to end the broadcasting career of Dan Rather, and that was a deliberate phishing attempt to end his career. The fact that "deepfake" had entered the lexicon long before anyone had seen one in the wild largely immunized all but the credulous against their effects, and the credulous are always only looking for an excuse to believe what they want to believe. I think the initial impression around LLMs has become "there's no reason to believe it" and it takes a lot longer to regain trust than it does to earn it in the first place. I suspect those who went balls-deep on ChatGPT will be regarded a year or so from now the same way those who went balls-deep on Google Glass: naive backers of an obviously corner-case idea who disregarded all notions of practicality. It's a tool. It has uses. Those uses are narrower than the boosters believe. The world shall continue in its orbit.
For the most part, I think you're correct. My main worry is that we'll find out that there's a lot of things that nobody in power really cares about being doing 'right,' especially when being done unreliably is essentially free. So things that were done with some amount of care and love purely because a human was doing them will be replaced with an AI that's 60% as good and 5% the cost and the world will be more mediocre than it was before